W3C HCLS – Clinical Decision Support

13 Sep 2012

See also: IRC log




<alejandrorg> hello, I can't join the telco via phone, I'm Alejandro Rodríguez

<matthias_samwald1> hello alejandro, thanks for the notice -- maybe in the next call!

<matthias_samwald1> i don't expect too many participants today, so the teleconference might end soon.

<matthias_samwald1> (i.e., there will not be much to see in this IRC log)

<matthias_samwald1> Amnon Shabo

[discussion of FHIR]

matthias_samwald1: met with Amnon Shabo, who's been active in the HL7 clinical genomics WG
... came up when discussing useing HL7 clinical genomics for this CDS task force
... it seems they tried to use genomics in CDA in pilot projects but it was overly complex
... he discussed creating a group at IHE for clinical genomics
... looking for people who could formally support such an IHE group
... i told him about what we're doing in HCLS and that we'd be interested in any overlap or cooperation points

ericP: would in IHE increase or decrease mindshare on our work?

matthias_samwald1: unclear now
... IHEs are tech-agnostic, just trying to fit stuff together
... W3C RECs or tech could be a part of an IHE
... i'll be contacted if anything happens [on this IHE project]
... i updated the ontological reprepresentation of alleles and haplotypes

<michel> excel

matthias_samwald1: doing this in excel is no untenable

<matthias_samwald1> https://drive.google.com/#folders/0ByGT-vnkGcoLazlNX2oxMUJzY3c

matthias_samwald1: (was easy to share with medical professionals)
... but now can't generate the OWL in excel
... instead using a php script to read the excel and spit out OWL

ericP: scale also constrains reasoners?

matthias_samwald1: yes
... currently the OWL doesn't load 'cause there's a mismatch between the a SNP in DBSNP and PharmGKB
... already found several inconsistencies between DBSNP and PharmGKB
... would expect that tr-owl will handle the scale
... note that it not complete and this incompleteness hides useful inferences
... will be working on this over the next few days
... trying to work with the entire PharmGKB haplotype table
... also the CPIC dosing guidelines
... if we can get this working with incomplete reasoning, we can work on completeness and in a realistic CDS timeframe

michel: intrested in inconsistencies between DBSNP and PharmGKB

matthias_samwald1: PharmGKB says that variant 123 can have any variant while DBSNP only permits the feasible variants
... plan to normalize. if PharmGKB uses random strands, we should normalize to DBSNP

<matthias_samwald1> rs1045642_G

matthias_samwald1, rs1045642_G comes from PharmGKB

scribe: entering that in DBSNP.

<michel> http://www.pharmgkb.org/rsid/rs1045642

scribe: in DBSNP, you can only have rs1045642_{A,C,T}

<michel> http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=1045642

matthias_samwald1: this is frequent, but not universal
... bobF wanted to represent the CPIC rules in a machine-processable format
... the lack of a formalism for expressing variants inspires me to position this formalism as a standard
... it would be good for HCLS to produce a terminology and a set of tools

michel: we started our work just representing this stuff in RDF/OWL, but your use of OWL to find inconsistencies is a welcome message

ericP: Note sufficient?

matthias_samwald1: SIOC was successful
... but would be cool to go beyond that
... pharmacogenomics is arguably the most important part of personalize medicine

ericP: could start with a Note and motivating the community to form a WG

michel: we can follow the history of provenance
... can start with the Note

matthias_samwald1: still plenty to do with finishing the prototype and then working on a scientific paper


michel: i wrote a parser for DBSNP
... would like to build up a mashup db
... planning to do a bio2rdf release this WE
... DBSNP not part of this release, but Oct 15 should include bigger databases
... we have code that RDF-izes clinicaltrials.org
... Rich Boyce has RDf of Structured Product Labels
... hopefully we'll have this in a reliable endpoint
... if you have a priority list of SNPs...

matthias_samwald1: my DBSNP converter maps a bit of the data in our OWL ontology

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2012/09/13 15:54:50 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

No ScribeNick specified.  Guessing ScribeNick: ericP
Inferring Scribes: ericP

WARNING: No "Present: ... " found!
Possibly Present: alejandrorg egonw ericP https matthias_samwald1 michel
You can indicate people for the Present list like this:
        <dbooth> Present: dbooth jonathan mary
        <dbooth> Present+ amy

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Got date from IRC log name: 13 Sep 2012
Guessing minutes URL: http://www.w3.org/2012/09/13-hcls-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]